An individualized transcranial electrical stimulation system based on a multi-scale fusion brain model and a method thereof

Abstract

The application discloses an individualized transcranial electrical stimulation system and method based on a multi-scale fusion brain model, which comprises the following modules: a large-scale brain model construction module, which is used for reconstructing an individualized brain function connection matrix by using electroencephalogram (EEG) data, then constructing an individualized large-scale brain model based on an average field model, and equivalent transcranial electrical stimulation to an external current input of the large-scale brain model, and constructing a large-scale brain model under the action of external stimulation; a microscopic neural circuit dynamics model construction module, which is used for fusing single neuron dynamics and synaptic plasticity, adopting a coupled leaky integrate-and-fire model and a spike-timing-dependent plasticity model to establish long-term plasticity between synapses, and then establishing a local neural circuit model through a stimulation target and downstream brain areas; an individualized multi-scale fusion brain model construction module, which is used for replacing corresponding nodes in the large-scale brain model with the local neural circuit model, and constructing an individualized multi-scale fusion brain model; and a stimulation response simulation module, which is used for adjusting transcranial electrical stimulation parameters by using the constructed individualized multi-scale fusion brain model, and obtaining EEG activity and short-term response and long-term effect. The application can more accurately describe the complexity of brain activity by establishing an individualized multi-scale fusion brain model.

Description

Technical Field

[0001] This invention belongs to the field of neuromodulation technology, specifically relating to a personalized transcranial electrical stimulation system and method based on a multi-scale fusion brain-like model. Background Technology

[0002] Transcranial electrical stimulation (tES) is a non-invasive technique that uses direct current (DC) or alternating current (AC) to modulate neural activity in the cerebral cortex. This technique has been widely applied in cognitive function modulation and neuropathological rehabilitation, demonstrating significant potential. Currently, personalized tES research primarily employs two approaches: one involves establishing a personalized head model based on individualized brain anatomy and then studying the electric field effects of tES through finite element analysis. However, this method neglects the brain's dynamic system characteristics and fails to capture the dynamic changes in brain activity before and after stimulation. The other approach involves constructing brain-like models based on anatomical or functional structures to study the brain's dynamic response to stimulation. However, these methods typically use a single, large-scale neural cluster model, considering only the immediate effects of electrical stimulation while neglecting the dynamic plasticity of brain structure and function—changes that continue to play a crucial role after stimulation ends. Therefore, there is an urgent need to develop a personalized tES method based on a multi-scale fusion brain-like model to more comprehensively and accurately understand the impact of tES on brain activity and provide stronger support for the development of personalized tES systems and methods. Summary of the Invention

[0003] To address the challenges of existing technologies, this invention provides a personalized transcranial electrical stimulation system and method based on a multi-scale fusion brain-like model. This invention improves the accuracy and efficiency of simulating brain network dynamics by using local neural circuit models to replace corresponding nodes in large-scale brain-like models and constructing personalized multi-scale fusion brain-like models.

[0004] To address the problems of the existing technology, the present invention adopts the following technical solution:

[0005] A personalized transcranial electrical stimulation system based on a multi-scale fusion brain-like model includes:

[0006] Large-scale brain-like model construction module: It is used to reconstruct an individualized brain functional connectivity matrix using EEG data, and then construct an individualized large-scale brain-like model based on the mean field model. Transcranial electrical stimulation is equivalent to the external current input of the large-scale brain-like model, and a large-scale brain-like model under the action of external stimulation is constructed.

[0007] Microscale neural circuit dynamics model construction module: It integrates single neuron dynamics and synaptic plasticity, uses a coupled leakage integral discharge model and a discharge time-dependent plasticity model to establish long-term plasticity between synapses, and then establishes local neural circuit models at the microscale through stimulation target points and downstream brain regions.

[0008] Personalized multi-scale fusion brain-like model construction module: Local neural circuit models are used to replace the corresponding nodes in the large-scale brain-like model to construct a personalized multi-scale fusion brain-like model;

[0009] Stimulation response simulation module: By using the constructed individualized multi-scale fusion brain-like model, the transcranial electrical stimulation parameters are adjusted to obtain the immediate response and long-term effects of brain electrical activity.

[0010] Furthermore, the large-scale brain-like model construction module is as follows:

[0011]

[0012] Where N is the number of EEG channels; E n (t) and I n (t) represent the average discharge rates of the excitatory and inhibitory modules in channel n, respectively, τ n It is the subgroup time constant in channel n. These represent the coupling strengths of excitatory-to-excitatory, excitatory-to-inhibitory, inhibitory-to-excitatory, and inhibitory-to-inhibitory modules in channel n, respectively; F represents the current activation activity; γ represents the activation gain; and θ represents the activation threshold. and These represent the external input currents of the excitatory and inhibitory clusters in channel n, respectively.

[0013] Furthermore, the local neural circuit model is as follows:

[0014]

[0015]

[0016] Where k = Exc and Inh represent excitatory neurons and inhibitory neurons, respectively. and V represents either an inhibitory neuron or an excitatory neuron, respectively. i k (t) represents the membrane potential of neuron i of type k; I noise For background noise, I ext For external input current; G represents the firing time of neuron j for the nth firing. j→i S represents the strength of the synaptic connection between presynaptic neuron j and postsynaptic neuron i; GABAS AMPA and S NMDA The synaptic dynamics of GABA receptors, AMPA receptors, and NMDA receptors are represented respectively; the synaptic dynamics are as follows:

[0017]

[0018] Where: α = AMPA, NMDA, GABA represent different receptors; τ i This indicates the firing time of the presynaptic neuron; For characteristic decay time constant and The rise time constants all depend on the receptor type; Θ(t) is the unit step function.

[0019] The present invention can also adopt the following technical solutions:

[0020] A method for individualized transcranial electrical stimulation using a multi-scale fusion brain-like model includes the following steps:

[0021] Step 1: Reconstruct an individualized brain functional connectivity matrix based on resting-state EEG data;

[0022] Step 2: Construct a personalized large-scale brain-like model based on the correlation between the mean-field model and the individualized brain functional connectivity matrix;

[0023] Step 3: Construct a large-scale brain-like model under external stimulation by using transcranial electrical stimulation as an equivalent external current input for an individualized large-scale brain-like model.

[0024] Step 4: Construct a microscale local neural circuit model by integrating single neuron dynamics and synaptic plasticity;

[0025] Step 5: Couple the large-scale brain-like model with the local circuit model to construct an individualized multi-scale fusion brain-like model;

[0026] Step Six: Using the individualized multi-scale fusion brain-like model, adjust the transcranial electrical stimulation parameters to obtain the immediate response and long-term effects of brain electrical activity.

[0027] The process of reconstructing an individualized brain functional connectivity matrix based on resting-state EEG data in step one includes:

[0028] Calculate the cross spectrum between any two channel signals x(t) and y(t).

[0029] S = X * (f)Y(f)

[0030] Where X(f) and Y(f) are the Fourier transforms of the two-channel signals x(t) and y(t), respectively, and * denotes complex conjugate;

[0031] The weighted phase hysteresis index between the two signals is calculated using the following formula;

[0032]

[0033] Where: S is the cross spectrum between signals. It is the imaginary part of S;

[0034] The calculated weighted phase lag indices are used to form a matrix to establish the EEG functional connectivity matrix.

[0035] Furthermore, the process of constructing a microscale local neural circuit model based on the fusion of single neuron dynamics and synaptic plasticity in step four includes: selecting target stimulation points by comparing the dynamic changes of each node before and during stimulation to screen the downstream brain regions indirectly affected by the stimulation based on the large-scale brain-like model under the action of external stimulation.

[0036] A coupled-leakage integral discharge model was used to construct local neural circuit neurons at the microscale of the stimulation target and downstream brain regions. These local neural circuit neurons consisted of excitatory neurons and inhibitory neurons; wherein:

[0037]

[0038] Where k = Exc and Inh represent excitatory neurons and inhibitory neurons, respectively. and V represents either an inhibitory neuron or an excitatory neuron, respectively. i k (t) represents the membrane potential of neuron i of type k; I noise For background noise, I ext For external input current; This represents the firing time of neuron j during its nth firing.

[0039] Adjust the parameters of the local neural circuit model based on the dynamic characteristics of the mean field model at the corresponding lead;

[0040] The strength of synaptic connections between neurons in a local neural circuit model can be adjusted using the following formula;

[0041]

[0042] Where j represents the presynaptic neuron; i represents the postsynaptic neuron;

[0043] The synaptic dynamics of different receptors in the local neural circuit model are established using the following formula;

[0044]

[0045] Where: α = AMPA, NMDA, GABA represent different receptors; τ i This indicates the firing time of the presynaptic neuron; For characteristic decay time constant and The rise time constants are all dependent on the receptor type; Θ(t) is the unit step function.

[0046] The long-term plasticity between synapses of excitatory neurons in a local neural circuit model is modulated using the following formula;

[0047]

[0048] Where Δg represents the change in the synaptic weight of the excitatory neuron, and the degree of excitatory synaptic regulation is affected by the regulation rate. and Restrictions on the value, time constant and These determine the length of the time windows for the enhancement and deactivation of excitatory neuronal synapses, respectively.

[0049] The long-term plasticity between inhibitory neuron synapses in a local neural circuit model is modulated using the following formula;

[0050]

[0051] Where Δg represents the change in the synaptic weight of the inhibitory neuron. Indicates the regulation rate; The time window is fixed; Δt represents the interval between presynaptic and postsynaptic discharges.

[0052] Beneficial effects

[0053] Compared with traditional technical solutions, the beneficial effects of this invention are:

[0054] 1. Traditional brain-inspired models typically consider only single-scale neural activity, while brain activity is multi-scale. The electrical activity of local neural circuits can propagate through brain networks to distant brain regions, thus affecting the dynamics of the whole-brain network. Therefore, this invention proposes to establish an individualized multi-scale fusion brain-inspired model. By combining local microcircuits and multi-scale fusion methods, the complexity of brain activity can be described more accurately.

[0055] 2. The multi-scale fusion brain-like model proposed in this invention not only considers single-neuron dynamics but also synaptic plasticity. This model can simulate not only the immediate response to transcranial electrical stimulation but also the long-term effects of electrical stimulation, thus more comprehensively considering the dynamic changes in brain activity.

[0056] 3. The model proposed in this invention adopts a local microscopic neural circuit model, which helps to reduce the computational burden of the algorithm and improve the simulation accuracy and efficiency of brain network dynamics. Attached Figure Description

[0057] Appendix Figure 1 A flowchart illustrating a personalized transcranial electrical stimulation method based on a multi-scale fusion brain-like model provided in an embodiment of the present invention;

[0058] Appendix Figure 2 This is a schematic diagram of a module of a personalized transcranial electrical stimulation system based on a multi-scale fusion brain-like model, provided as an embodiment of the present invention. Detailed Implementation

[0059] The following is in conjunction with the appendix Figure 1-2 The present invention is described as follows:

[0060] like Figure 1 As shown, this invention also discloses a personalized transcranial electrical stimulation system based on a multi-scale brain-like model, comprising a large-scale brain-like model construction module, a micro-scale neural circuit dynamics model construction module, a multi-scale fusion brain-like model construction module, and a stimulus response simulation module. It includes:

[0061] Step 101: Reconstruct an individualized brain functional connectivity matrix based on resting-state EEG data to build an individualized large-scale brain-like model;

[0062] Step 2.102: Transcranial electrical stimulation is equivalent to the external current input of a large-scale brain network dynamics model, and a large-scale brain-like model under the action of external stimulation is constructed.

[0063] Step 3.103: Integrate single neuron dynamics and synaptic plasticity to construct a microscale local neural circuit model;

[0064] Step 4.104: Couple the large-scale brain network dynamics model with the local circuit model to construct an individualized multi-scale fusion brain-like model;

[0065] Step 5.105: Using the individualized multi-scale fusion brain-like model, adjust the transcranial electrical stimulation parameters to obtain the timely response and long-term effects of brain electrical activity.

[0066] Example 1

[0067] Please see Figure 1 , Figure 1 This is a schematic flowchart of a personalized transcranial electrical stimulation (TCS) method based on a multi-scale fusion brain-like model, provided by an embodiment of the present invention. The personalized TCS method based on a multi-scale fusion brain network dynamics model provided by this embodiment includes steps S1 to S5, wherein:

[0068] S1: Acquire resting-state EEG data, and establish an individualized large-scale brain-like model based on the EEG data, specifically:

[0069] S1.1 Based on resting-state EEG data, the weighted phase lag index between any two channels is calculated to represent the degree of synchronization between the two channels. The calculated weighted phase lag indices are then combined into a matrix to obtain the EEG functional connectivity matrix, specifically:

[0070] Calculate the cross spectrum between any two channel signals x(t) and y(t).

[0071] S = X * (f)Y(f)

[0072] Here, X(f) and Y(f) are the Fourier transforms of the two-channel signals x(t) and y(t), respectively.

[0073] Calculate the weighted phase hysteresis exponent between the two signals mentioned above.

[0074]

[0075] Where S is the cross spectrum between signals. It is the imaginary part of S.

[0076] The calculated weighted phase lag exponents are combined into a matrix to obtain the EEG functional connectivity matrix.

[0077] S1.2 A mean-field (WC) model is used to simulate the dynamic characteristics of each channel. Each WC model contains one excitatory cluster and one inhibitory cluster. The excitatory cluster is used to characterize the firing rate of the excitatory neuron cluster, and the inhibitory cluster is used to characterize the firing rate of the inhibitory neuron cluster. Specifically:

[0078]

[0079] Where N is the number of EEG channels. Where E n (t) and I n (t) represent the average firing rates of neurons in the excitatory and inhibitory modules of channel n, respectively. n It is the subgroup time constant in channel n. These represent the coupling strengths of excitatory-to-excitatory, excitatory-to-inhibitory, inhibitory-to-excitatory, and inhibitory-to-inhibitory modules in channel n, respectively, with F representing the current activation activity.

[0080] S1.3 uses the method in S1.1 to calculate the functional connectivity matrix of the large-scale brain-like model, maximizes the correlation between the model's functional connectivity matrix and the EEG data's functional connectivity matrix, and obtains an individualized large-scale brain-like model.

[0081] S2: Based on the individualized large-scale brain-like model established in S1, a large-scale brain-like model under external stimuli is constructed, specifically as follows:

[0082] S2.1 equates transcranial electrical stimulation to the external current input of a large-scale brain-like model, constructing a large-scale brain-like model under external stimulation, specifically as follows:

[0083]

[0084] in, and These are the external input currents for the excitatory and inhibitory clusters in channel n, respectively.

[0085] S2.2 Based on the large-scale brain-like model under external stimulation established in S2.1, target stimulation points are selected, and the dynamic changes of each node in the large-scale brain-like model before and during stimulation under external stimulation are compared, including oscillation characteristics and the degree of synchronization between nodes, to screen out the downstream brain regions indirectly affected by the stimulation.

[0086] S3: Integrating single-neuron dynamics and synaptic plasticity, a microscale local neural circuit model is constructed, specifically:

[0087] S3.1 A coupled leak-integrated discharge (LIF) model was used to construct local neural circuit neurons at the microscale of the stimulation target and downstream brain regions. Each circuit model contained M neurons, with a ratio of excitatory to inhibitory neurons of 4:1. The parameters of the local neural circuit model were adjusted according to the mean-field model dynamics characteristics at the corresponding leads, specifically:

[0088]

[0089] Where k = Exc and Inh represent excitatory neurons and inhibitory neurons, respectively. and V represents either an inhibitory neuron or an excitatory neuron, respectively. i k (t) represents the membrane potential of neuron i of type k; G represents the firing time of neuron j for the nth firing. j→i This represents the strength of the synaptic connection between presynaptic neuron j and postsynaptic neuron i; and the initial values ​​are set as follows:

[0090]

[0091] S α (t) represents the synaptic dynamics of different receptors α = AMPA, NMDA, and GABA, respectively, using a double exponential function, where the characteristic decay time constant is... and rise time constant Both depend on the receptor type. τ i This indicates the firing time of the presynaptic neuron. S α The form of (t) is as follows:

[0092]

[0093] Where Θ(t) is the unit heaviside step function.

[0094] S3.2 employs a spike-timing-dependent plasticity model to simulate long-term plasticity between synapses, mimicking the long-term effects following external stimuli. Specifically:

[0095] For the excitatory discharge time-dependent plasticity model,

[0096]

[0097] For the time-dependent plasticity model of suppressive discharge

[0098]

[0099] Where Δg represents the change in synaptic weight, and the degree of excitatory synaptic regulation is affected by the regulation rate. and Restrictions on the value, time constant and These factors determine the time window lengths for excitatory synaptic enhancement and deactivation, respectively. Inhibitory synaptic enhancement, on the other hand, is primarily determined by the regulation rate. and time window Decision. Δt represents the interval between presynaptic and postsynaptic discharges, defined as:

[0100] Δt=t post -t pre

[0101] Among them, t pre and t post These represent the presynaptic and postsynaptic discharge times, respectively.

[0102] S4: Couple large-scale brain-like models with micro-scale local circuit models to construct individualized multi-scale fusion brain network dynamics models.

[0103] S4.1 uses a local loop model to replace the corresponding nodes in the large-scale brain-like model to construct an individualized multi-scale fusion brain-like model.

[0104] S5: Using the individualized multi-scale fusion brain-like model constructed in S4, transcranial electrical stimulation parameters were adjusted to obtain the immediate response and long-term effects of brain electrical activity, specifically:

[0105] S5.1 Based on the given transcranial electrical stimulation parameters, adjust the external current input of the model to obtain the timely response of brain electrical activity under electrical stimulation.

[0106] S5.2 After removing external stimuli, the spontaneous evolution of the multi-scale brain network dynamics model is obtained, simulating the long-term effects after electrical stimulation.

[0107] Example 2

[0108] Please see Figure 2 , Figure 2 This is a schematic diagram of a module of a personalized transcranial electrical stimulation system based on a multi-scale fusion brain-like model, provided as another embodiment of the present invention. The personalized transcranial electrical stimulation system based on a multi-scale fusion brain-like model, provided as another embodiment of the present invention, includes:

[0109] Large-scale brain-like model construction module: It is used to reconstruct an individualized brain functional connectivity matrix using EEG data, and then construct an individualized large-scale brain-like model based on the Wilson-Cowan model. Transcranial electrical stimulation is equivalent to the external current input of the large-scale brain-like model, and a large-scale brain-like model under the action of external stimulation is constructed.

[0110] Microscale neural circuit dynamics model construction module: Integrating single neuron dynamics and synaptic plasticity, it uses the leaky integrate-and-fire (LIF) model and the firing time-dependent plasticity model to simulate the long-term plasticity between synapses, and constructs local neural circuit models at the microscale of stimulation targets and downstream brain regions.

[0111] Multi-scale fusion brain-like model construction module: Local loop models are used to replace the corresponding nodes in the large-scale brain-like model to construct an individualized multi-scale fusion brain-like model;

[0112] Stimulation response simulation module: By using a constructed individualized multi-scale fusion brain-like model, the transcranial electrical stimulation parameters are adjusted to obtain the immediate response and long-term effects of brain electrical activity.

[0113] Although the present invention has been described above, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many modifications under the guidance of the present invention without departing from the spirit of the present invention, and these modifications are all within the protection scope of the present invention.

Claims

1. A personalized transcranial electrical stimulation system based on a multi-scale fusion brain-like model, characterized in that, include: Large-scale brain-like model construction module: It is used to reconstruct an individualized brain functional connectivity matrix using EEG data, and then construct an individualized large-scale brain-like model based on the mean field model. Transcranial electrical stimulation is equivalent to the external current input of the large-scale brain-like model, and a large-scale brain-like model under the action of external stimulation is constructed. Microscale neural circuit dynamics model construction module: It integrates single neuron dynamics and synaptic plasticity, uses a coupled leakage integral discharge model and a discharge time-dependent plasticity model to establish long-term plasticity between synapses, and then establishes local neural circuit models at the microscale through stimulation target points and downstream brain regions. Personalized multi-scale fusion brain-like model construction module: Local neural circuit models are used to replace the corresponding nodes in the large-scale brain-like model to construct a personalized multi-scale fusion brain-like model; Stimulation response simulation module: By using the constructed individualized multi-scale fusion brain-like model, the transcranial electrical stimulation parameters are adjusted to obtain the immediate response and long-term effects of brain electrical activity.

2. The personalized transcranial electrical stimulation system based on a multi-scale fusion brain-like model according to claim 1, characterized in that, The large-scale brain-like model construction module is as follows: Where N is the number of EEG channels; E n (t) and I n (t) represent the average discharge rates of the excitatory and inhibitory modules in channel n, respectively, τ n It is the subgroup time constant in channel n. These represent the coupling strengths of excitatory-to-excitatory, excitatory-to-inhibitory, inhibitory-to-excitatory, and inhibitory-to-inhibitory modules in channel n, respectively; F represents the current activation activity; γ represents the activation gain; and θ represents the activation threshold. and These represent the external input currents of the excitatory and inhibitory clusters in channel n, respectively.

3. The personalized transcranial electrical stimulation system based on a multi-scale fusion brain-like model according to claim 1, characterized in that, The local neural circuit model is as follows: Where k = Exc and Inh represent excitatory neurons and inhibitory neurons, respectively. and V represents either an inhibitory neuron or an excitatory neuron, respectively. i k (t) represents the membrane potential of neuron i of type k; I noise For background noise, I ext For external input current; G represents the firing time of neuron j for the nth firing. j→i S represents the strength of the synaptic connection between presynaptic neuron j and postsynaptic neuron i; GABA S AMPA and S NMDA The synaptic dynamics of GABA receptors, AMPA receptors, and NMDA receptors are represented respectively; the synaptic dynamics are as follows: Where: α = AMPA, NMDA, GABA represent different receptors; τ i This indicates the firing time of the presynaptic neuron; For characteristic decay time constant and The rise time constants are all dependent on the receptor type; Θ(t) is the unit step function.

4. A method for individualized transcranial electrical stimulation using the multi-scale fusion brain-like model as described in claim 1, characterized in that, Includes the following steps: Step 1: Reconstruct an individualized brain functional connectivity matrix based on resting-state EEG data; Step 2: Construct a personalized large-scale brain-like model based on the correlation between the mean-field model and the individualized brain functional connectivity matrix; Step 3: Construct a large-scale brain-like model under external stimulation by using transcranial electrical stimulation as an equivalent external current input for an individualized large-scale brain-like model. Step 4: Construct a microscale local neural circuit model by integrating single neuron dynamics and synaptic plasticity; Step 5: Couple the large-scale brain-like model with the local circuit model to construct an individualized multi-scale fusion brain-like model; Step Six: Using the individualized multi-scale fusion brain-like model, adjust the transcranial electrical stimulation parameters to obtain the immediate response and long-term effects of brain electrical activity.

5. The method for individualized transcranial electrical stimulation using the individualized multi-scale fusion brain-like model according to claim 4, characterized in that, The process of reconstructing an individualized brain functional connectivity matrix based on resting-state EEG data in step one includes: Calculate the cross spectrum between any two channel signals x(t) and y(t). S=X * (f)Y(f) Where X(f) and Y(f) are the Fourier transforms of the two-channel signals x(t) and y(t), respectively, and * denotes complex conjugate; The weighted phase hysteresis index between the two signals is calculated using the following formula; Where: S is the cross spectrum between signals. It is the imaginary part of S; The calculated weighted phase lag indices are used to form a matrix to establish the EEG functional connectivity matrix.

6. The method for individualized transcranial electrical stimulation using the individualized multi-scale fusion brain-like model according to claim 5, characterized in that, Step four involves constructing a microscale local neural circuit model based on the fusion of single-neuron dynamics and synaptic plasticity; it includes: Based on the large-scale brain-like model under the action of external stimulation, the downstream brain regions indirectly affected by the stimulation are screened by selecting target stimulation points and comparing the dynamic changes of each node before and during stimulation. A coupled-leakage integral discharge model was used to construct local neural circuit neurons at the microscale of the stimulation target and downstream brain regions. These local neural circuit neurons consisted of excitatory neurons and inhibitory neurons; wherein: Where k = E and I represent excitatory neurons and inhibitory neurons, respectively. V represents the y-th neuron connected to neuron x. i k (t) represents the membrane potential of neuron i of type k; I noise For background noise, I ext For external input current; This represents the firing time of neuron j during its nth firing. Adjust the parameters of the local neural circuit model based on the dynamic characteristics of the mean field model at the corresponding lead; The strength of synaptic connections between neurons in a local neural circuit model can be adjusted using the following formula; Where j represents the presynaptic neuron; i represents the postsynaptic neuron; The synaptic dynamics of different receptors in the local neural circuit model are established using the following formula; Where: α = AMPA, NMDA, GABA represent different receptors; τ i This indicates the firing time of the presynaptic neuron; For characteristic decay time constant and The rise time constants are all dependent on the receptor type; Θ(t) is the unit step function; The long-term plasticity between synapses of excitatory neurons in a local neural circuit model is modulated using the following formula; Where Δg represents the change in the synaptic weight of the excitatory neuron, and the degree of excitatory synaptic regulation is affected by the regulation rate. and Restrictions on the value, time constant and These determine the length of the time window for the enhancement and deactivation of excitatory neuronal synapses, respectively; Δt represents the interval between presynaptic and postsynaptic firing. The long-term plasticity between inhibitory neuron synapses in a local neural circuit model is modulated using the following formula; Where Δg represents the change in the synaptic weight of the inhibitory neuron. Indicates the regulation rate; The time window is fixed; Δt represents the interval between presynaptic and postsynaptic discharges.

Images